I always enjoy a challenge when it comes to doing something in PSpice.
From Rose's blog, post #58:
Dear Reader
I am not sure that the waveform can be duplicated on a simulator. I'd be interested to see if this is, in fact, possible. Perhaps our Poynty can give this a go. I am delighted, in any event, to see that it's being denied on the basis of a faulty IRFPG50.
Indeed I can, and I have... for the most part. And you may want to pay keen attention to my findings.
There are 3 differences in the wave forms however, which I will explain below.
osc_repl01.png is greatly representative of the "phenomenon" that Rose is seeing in her wave forms. Specifically, this is in regards to the high amplitude oscillation present on the battery line (and Drain) and shunt measurement point during the period of time in the switch cycle where the MOSFET switch is "OFF".
The
difference #1 one will observe here between this and Rose's wave forms, is that the ringing is damped (i.e. falls off in amplitude with time) with my wave forms, while in the wave forms Rose posted, the amplitude appears quite constant.
The cause for the constant amplitude ringing in the wave forms Rose posted, is that the MOSFET VGS and VD are in phase. VGS is being "modulated" by the "noisy" signal on the Source pin, caused by the high impedance path back to the battery. This has the effect of applying negative feedback, and since the MOSFET is a voltage controlled resistance, the D-S resistance is also being modulated in such a fashion to "compress" the Drain voltage. Think of an AGC type control.
osc_repl02.png is a zoomed-in portion of the ringing.
Difference #2 is that the phase shift in my wave forms is 90 degrees, while Rose's show 180 degrees. This clearly indicates that there is a second order network in the actual circuit.
I have not included all the parasitic components around the MOSFET, so this may account for the single-order phase shift, vs. second order. With more time, I am fairly confident I could recreate this effect also, however, for the purposes of this post it is not required, as will soon become evident.
The
#3 difference is simply the slow ramp-ON and ramp-OFF of oscillation shown in Rose's wave forms, vs. the sharp ON and OFF transitions seen in my simulation wave forms. For the simulation, I have used an ideal switch, and so it does not exhibit this ramp effect. It is interesting to note that Rose's wave forms clearly indicate that the circuit exhibits ringing, even when the switch is ON.
Here is where you will want to pay particular attention Rose:There are primarily two causes for your present circuit to exhibit this unusual ringing effect:
1) The presence of an excessive battery cable length. This introduces a relatively large amount of inductance into the circuit, and effectively causes the battery to "float" relative to where you are taking your measurement points, i.e. near the MOSFET (see the attached photo of Rose's setup).
2) The presence of a completely blown, or severely damaged MOSFET body diode. With the low impedance heater coil and much higher supply voltage than before (24V in previous tests, now about 48V), the MOSFET body diode is experiencing repeated high energy avalanche, to the point it has evidently caused damage internal to the MOSFET currently installed in your circuit. I am 90% certain, that if you were to replace that MOSFET with a known good IRFPG50, you would see some very different wave forms, at least until that one sustains damage as well.
So there you have it Rose. I strongly encourage you to replace your MOSFET and see for yourself. With a proper working body diode, those lengthy oscillations have no chance to persist as we see in your wave forms.
.99
"Some scientists claim that hydrogen, because it is so plentiful, is the basic building block of the universe. I dispute that. I say there is more stupidity than hydrogen, and that is the basic building block of the universe." Frank Zappa